In US practice, the Genealogical Proof Standard replaced the older “Preponderance of Evidence” rule in the late 20th century to distance genealogy from a purely legal definition. However, the GPS is essentially “Preponderance of Evidence” in a more expensive suit. While the Board for Certification of Genealogists (BCG) emphasizes that a “preponderance” (simply being 51% sure) is insufficient for formal proof, the GPS still relies on the researcher’s subjective assessment of “the weight of the evidence.” When you “resolve a conflict” between a death certificate and a family Bible, you are inherently deciding which piece of evidence is more reliable.
The GPS fails to escape the shadow of “preponderance” because “proof” in genealogy is never absolute; just a qualitative statement of probability based on a “reasonably exhaustive search.” By requiring a written conclusion that explains away contradictions, the GPS forces a researcher to build a narrative where one set of facts “outweighs” another. However, by claiming to be more rigorous than a “preponderance,” the GPS creates a false sense of security. If you follow the five steps, you have met the standard. What that means in reality is you have simply reached the most persuasive argument allowed by the surviving (and sometimes biased) documentation.
A better approach would be to require, and state in your conclusion, a quantitative probability estimate. I wrote about this approach, with a case study, a decade ago in the Spring 2016 issue of BIFHSGO’s Anglo-Celtic Roots in Confidence and the Genealogical Proof Standard.
To be statistically correct, a probability estimation is base on a number of fundamental assumptions which are hard to apply to genealogical data. Without meeting these assumptions a statement of probability is a guess, which is hopefully informed.